Silicon Electrochemistry in Molten Salts.

[NNadir]

Last edited Sun Mar 8, 2020, 06:34 AM - Edit history (1)

The paper I will discuss in this post is a relatively short review article, this one: Silicon Electrochemistry in Molten Salts (Eimutis Juzeliu̅nas and Derek J. Fray*, Chemical Reviews 2020 120 (3), 1690-1709).

Derek Fray is one of the "F's" in the FFC Cambridge process, invented by Derek Fray and Tom Farthing, George Chen, a process which has the potential to change the processing of titanium, making metallic titanium readily available much as the Hall-Heroult process made metallic aluminum available.

This nice periodic table, to which I Googled, shows the abundance of the elements in Earth's crust:



Both aluminum and titanium are common elements, but despite that fact, the difficulty of reducing their ores, bauxite (alumina) and ilumite, rutile and anastase, ores of titanium, to the metals long prevented them from being common for most of human history. It is only in the last 100 years, since the development of the Hall-Heroult process, that aluminum metal has become common.

The FFC Cambridge process has the potential to do for titanium - a metal lighter and stronger and with a higher melting point than most steel alloys - what Hall Heroult did for Aluminum, make it common.

These processes are electrochemical processes, but in many cases they are not purely electrical in nature, as both the FFC Cambridge process and the Hall-Heroult process rely on carbon anodes, which are oxidized in the process.

In the Hall-Heroult process, the molten salt is cryolite, a mineral once mined in Greenland, which was a major source, now depleted, and is now made synthetically by treating bauxite with hydrofluoric acid and neutralizing it with NaOH:

6 NaOH + Al2O3 + 12 HF ? 2 Na3AlF6 + 9 H2O. (This is actually a two-step process, obviously; first the alumina is dissolved in HF, and then it hexafluoroaluminato acid is neutralized with NaOH.)

The Hall-Heroult process is energy intensive, cryolite melts at 950 °C, and the melting point rises when alumina, Al2O3 is dissolved in it. It is also true, just as it is for power plants, that allowing aluminum torts to cool wastes energy; making them useless for variable energy, including the so called "renewable energy" on which we bet the planetary atmosphere and which failed to address the on going destruction of that atmosphere.

In this space I covered another reason that the production of aluminum is dependent of dangerous fossil fuels, besides the need for continuous operation: This is the nature of the anode, which is made from petroleum coke:

Can Biocoke Address the Anode CO2 Problem (Owing to Petroleum Coke) for Aluminum Production?

In that text, I mentioned that besides the dangerous fossil fuel waste carbon dioxide, another waste product is carbon tetrafluoride:

The paper does not mention another product of the "burn up" of carbon anodes, which is CF4, carbon tetrafluoride, as the Hall process for aluminum releases copious amounts of CF4 (perfluoromethane), which according to the 5th climate assessment report has a global warming potential, a measure of how much worse it is than carbon dioxide, of 6,630, a factor of more than 230 times greater than methane itself.

According to the aluminum industry,the mean fluoride intensity of aluminum production is 0.58 kg "F"/ton Al. Chemically, all of this "F" is released as CF4. The mean atmospheric lifetime of CF4 in the planetary atmosphere, where the main sink is radiation, is on the order of 50,000 years.

On some level the FFC process has the potential to be worse, since the molten salt in its most prevalent form is calcium chloride as opposed to a fluoride metallate salt like cryolite. While CF4 is a problematic gas, carbon tetrachloride is a fairly powerful toxin; it is actually used in research to destroy the livers of test animals while they are alive. Yes, it can do that to you too. (Many older chemists will be familiar with using CCl4; hopefully generally in the hood.)

One of the reasons I write posts here is that writing them inspires me to look things up, however, and since I've wondered about anodes for quite some time for the oxidation of highly electronegative elements, oxygen and fluorine, I looked into anodes for the FCC process and a number of interesting approaches were located in the literature by looking into references to FFC papers, many of which were authored by Derek Fray.

(No one should be surprised if and when Fray wins a Nobel Prize in chemistry.)

Here's a cool one, calcium titanate doped with calcium ruthenate: Development of an Inert Anode for Electrowinning in Calcium Chloride–Calcium Oxide Melts

(I had the pleasure of discussing this paper with my son late last night after picking him up for his spring break while we driving home. Life is full on indescribable joys and then you die.)

The FCC Cambridge process can be controlled by careful management of the conditions under which it operates to produce, with inert electrodes such as calcium ruthenate doped calcium titanate - there are many others - relatively pure oxygen as opposed to chlorine or (with carbon based electrodes, chlorocarbons.

(For the record, ruthenium is a relatively rare cogener of iron, however significant amounts of it can be isolated from used nuclear fuel, particularly in fuel cycles based on plutonium fission - my favorite kind of fission fuel. There is one isotope present in fresh used nuclear fuel, 106Ru that is radioactive, with a half-life of 373. days. However, given the fact that this isotope is diluted with the stable isotopes 100Ru, 101Ru, 102Ru, 104Ru and 105Ru, after 10 to 20 years the residual radioactivity from it, and 103Ru (half-life 39.26 days is trivial. There is good reason for isolating ruthenium from used nuclear fuel while it is still radioactive, since Ru-106 decays into valuable mono-isotopic palladium-106, and Ru-103 decays into even more valuable rhodium-103, rodium's only stable isotope. Ruthenium can be removed from used nuclear fuels by oxygen volatilization as the tetraoxide which can be distilled.)

It turns out that the FFC Cambridge process, although originally focused on titanium is useful for many other metals, including many that have historically been reduced using coal, for example, iron. The reaction is now generalized sometimes in the literature as the "electro-deoxygenation reaction." It has been utilized to reduce a wide array of metals, from tantalum, to zirconium to hafnium to neodymium etc...and relevant to the paper under discussion, the semi-metal silicon, which as we all know, is a key element in electronic devices including the solar cells on which many of on the left have proved willing to bet the planetary atmosphere, despite zero evidence that it has worked, is working or will work.

From the introduction to the paper:

Silicon (Si) lies at the heart of modern technology. It is widely used to produce metal alloys, silicon organic compounds, optical fibers, solar elements, advanced ceramics, batteries, microchips, and numerous other advantageous applications. The greatest part of Si production (ca. 50%) is used to produce alloys (e.g., Fe–Si, Al–Si, Al–Mg–Si, Cu–Si, Si–Mn, etc.). A low purity material, so-called metal-grade silicon (MeG-Si) with 98–99% of silicon, is used to produce alloys and organic compounds. For solar and electronic devices, there is a need for solar grade silicon (SoG-Si) with a purity of 99.9999% (6N) or electronics grade silicon (EG-Si) with a purity of 99.9999999% (9N).

Silicon electrochemistry in molten salts has recently attracted considerable attention due to its potential to produce SoG-Si with negligible carbon footprint. Advantages of the electrochemical methods include usage of electrons as clean reduction agents instead of harmful chemicals as well as its simplicity owing to one-step electro-deoxidation. Thus, silicon electrochemistry is of general interest and of current relevance because of its potential to significantly contribute to a low-carbon economy.
Many nations declared their support at the United Nations Climate Change Conference for the fight against climate change and keeping the global temperature rise in this century well below 2 °C (the Paris Climate Summit 2015, COP21). To achieve such goals, solar energy will be a future core energy source. Such global tasks would be achieved using technologies based on Earth-abundant materials.

Silicon, as the second most abundant element and being a nontoxic, efficient, and robust material, will dominate in the solar energy market for, at least, the next few decades.(1) Over 90% of solar cells are produced from silicon, where total wafer cost represents more than 60% of the overall cell costs. The main disadvantage of silicon, however, is its poor intrinsic ability to absorb light with more than 30% of incident light being reflected if the silicon surface is not specifically textured coated with antireflection coatings. Also, high processing complexity and considerable material loss when slicing Si ingots into wafers increases the cost of photovoltaic devices.

The "high processing" complexity involves the carbothermic reduction of silicon, using carbon derived from dangerous fossil fuels, the dissolution of crude elemental silicon with hydrochloric acid to make trichlorosilane (for the purpose of purification), rearrangement of the trichlorosilane to give silane, SiH4, and tetrachlorosilicon in, ideally, a 1 to 3 ratio. The silane is then reduced to polysilicon at high temperatures to give polysilicon for "green" solar cells. (Silicon tetrachloride can be hydrogenated temperatures to regenerate trichlorosilane.)

There is nothing environmentally friendly about this process.

So this brings us to the FFC Cambridge process for the electro-deoxygenation of silicon dioxide, sand.

The authors of the review article discuss the size of the silicon market and offer us a nice graphic about it:



The caption:

Figure 1. Silicon world production (2001–2018) in millions of metric tons (Mt). Source: Mineral commodity summaries, U.S. Department of the Interior, U.S. Geological Survey.

They remark on the carbon cost of producing this silicon.

Dynamics of Si world production (2001–2018) in Figure 1 shows that the market experienced remarkable growth from 2001 to 2011 and reached some saturation over the past years with around 7.0 millions of metric tons (Mt) per year. According to (2), such an amount of Si is equivalent to CO2 emission of 11.0 Mt per year. There are nondirect emissions as well, for instance, those related to heat generation. The reduction process itself is endothermic with energy consumption of about 10.5 MWh per one ton of Si.

While 11 million tons of carbon dioxide may sound like a lot - and clearly this figure ignores the process heat required to make polysilicon - it isn't. It's trivial. The world is currently releasing about 35 billion tons of carbon dioxide per year, mostly because the half a century of hoopla about so called "renewable energy" has not worked, is not working and will not work to address climate change.

As of the 2018, the last year for which we have reasonably complete figures, the much worshiped (to the detriment of all future generations) wind and solar industries combined produced slightly more than 2% of the world energy demand, with a fair fraction of that energy wasted in a Quixotic enterprise of attempting to store it in environmentally dubious batteries, this while driving up the operating costs of the necessary redundant dangerous fossil fuel systems required to back it up, by requiring, at a thermodynamic, economic and environmental cost, these systems to momentarily shut down so uneducated people with no interest in the engineering that drives their lives can cheer. Of this 2%, the overwhelming amount of it actually came from the wind industry and not the solar industry. If we're overly generous however and assume, with no support that 1% of the world's energy - half of 2% of useless wind and solar fantasy - about 6 exajoules out of 600 exajoules, we can estimate about how much silicon would be required, about one billion tons, releasing about 1.5 billion tons of carbon dioxide, not counting process heat, and not counting the material processing carbon costs of billions upon billions of batteries.

Yeah Green...

The authors remark on the additional carbon cost of silicon processing beyond the carbon generated in the carbothermic reduction of sand:

The carbothermic reduction gives a MeG-Si, which contains a high percentage of impurities (Si–98–99%, the rest – Fe, Al, B, P, Ca, Mg, Mn, Ti, C, etc.). SoG-Si is industrially produced from MeG-Si feedstock predominantly by the Siemens process. Si is converted using hydrochloric acid to volatile trichlorosilane at 350 °C (SiHCl3, the boiling point is Tb = 31.8 °C), which makes it possible to separate silicon from impurities by distillation in fractionating columns. Trichlorosilane is decomposed to Si on pure Si bars heated to around 1100 °C.

Carbothermic deoxidation and purification require a large amount of energy and emit CO2. The challenge is to extract Si with zero carbon footprint using electrons as absolutely clean reducing agents generated by sustainable renewable sources.

By the way, an explosion of a trichlorosilane reactor in Japan instanteously killed an infinitely larger amount of people who than were instantaneously killed by exposure to radiation from the much discussed destruction of the nuclear reactors at Fukushima, since the number of people instanteously killed by radiation was zero in the event, and the number of people killed in the trichlorosilane explosion on January 9, 2014 was five people, (injuring 12) and five is infinitely larger than zero.

But maybe the FFC silicon process can help, no?

Maybe, but the bulk of the world's electricity still comes from dangerous fossil fuels - the use of which is rapidly rising not falling - and the FFC Cambridge process involves lots and lots of heat to melt the calcium chloride salt.

Here is a table from the paper showing some literature in which electro-deoxygenation has been explored to reduce silicon dioxide to elemental silicon:



It does seem in every case in these papers, carbon was the anode: No calcium ruthenate doped calcium titanate was used, which is not to say it couldn't be used.

Anyway, I've prattled on too long about this interesting sidelight in the reality connected with disconnected dreams about saving the dying world with silicon.

Some pictures from the paper:





The caption:

Figure 2. Open circuit potential (Eocp) vs time curves obtained for Si in the Mo frame (Si–Mo), Si with a 2 ?m thermally coated SiO2 layer in the Mo frame (Si–SiO2–Mo), and a single Mo frame in molten CaCl2 at 850 °C. Potential has been measured against the graphite pseudoreference electrode. The samples were cut from the p-Si (100) wafer (S = 1.3 × 5.0 = 6.50 cm2). The SiO2 layer was thermally coated on the p-Si wafer. Mo wire (⌀ 0.5 mm) has been used as a single electrode or as a contacting one in the form of a frame attached to Si as shown in ref (20).

Here's some commentary in the text on figure 2:

In addition, the electrochemical behavior of the system Mo–Si–CaCl2, which is usually used in experimental configurations (Table1), will be discussed. The basic knowledge of the Si electrode helps to better identify the processes of technological significance. Open circuit potential (Eocp) characterizes electrochemical silicon activity: a negative shift of Eocp means an increase in electrode energy to release electrons or, in other words, an increase in its oxidation power.

Figure 2 shows variations in Eocp recorded for Si–Mo, Si–SiO2–Mo, and single Mo electrodes in molten CaCl2. The potential of single Mo electrode shifts during the exposure by ?Eocp ? 250 mV, which indicates passivation of the surface. Attachment of silicon shifts the curve down by ?Eocp ? ?0.5 V, which means the silicon potential is steady.

My son just informed me that his internship this summer will be in England, where he will be working on micro and nanostructured silicon carbides.

He seems to be spending a lot of time on SEM and related microscopy stuff these days. Some SEM images of silica pellets:



The caption:

Figure 3. SEM images of (a) the nano-SiO2 pellet sintered at 900 °C and (b) that after immersion in molten CaCl2 at 900 °C for 30 min. (c) The surface morphology of p-Si coated with a thermal 2 ?m SiO2 layer after 20 min exposure in molten CaCl2 at 850 °C. Images (a) and (b) reprinted with permission from ref (27). Copyright 2009 The Royal Society of Chemistry.

Current voltage curves:




The caption:

Figure 4. CV curves obtained for Mo (red) and Si–Mo (dark) electrodes obtained in molten CaCl2 at v = 100 mV s–1 and T = 850 °C. Potentials are given with respect to the equilibrium potential E0Ca/Ca2+. The first cycle was started at Eocp = 0.33 V for Si–Mo and Eocp = 0.85 V for Mo. The figure displays a third cycle, which shows a loop when the starting and the final points are coincident. The samples were cut from the p-Si (100) wafer (S = 1.3 × 5.0 = 6.50 cm2). Mo wire (⌀ 0.5 mm) has been used as a single electrode (red) or as a contacting electrode in the form of a frame attached to Si, as shown in ref (20).

Direct formation of semiconductor diodes in molten salt apparatus is shown in this cartoon:



The caption:

Figure 5. Electrochemical formation of the Si p-n junction: (a) molten salt cell and (b) outline of the two-step formation of n-type and p-type films. Reprinted with permission from ref (69). Copyright 2017 American Chemical Society.

There is some discussion of similar processes in molten fluoride salts, and then a discussion of using electrochemical methods to purify silicon avoiding all that rather nasty chlorosilane/silane chemistry:



The caption:

Figure 6. Principle of Si purification using the Al–Si liquid cathode, Cu–Si liquid anode, molten salt electrolyte, and impure Si feedstock. Reprinted with permission from ref (93). Copyright 2016 The Electrochemical Society.

The caption:

Figure 7. SEM images of the surface structures of electrochemically treated p-Si ⟨100⟩ in molten CaCl at 900 °C: (a) macroporous surface formed by anodizing at E = 1.9 VCa/Ca2+ for 5 min; (b) highly disordered Si surface obtained when applying five polarization cycles (?1.17 ? E ? 1.83 VCa/Ca2+); (c) cross section of the macroporous structure (d ? 10 ?m) formed on the Si substrate (upper part) by polarizing at E = ?0.07 VCa/Ca2+ for 3 h at T = 950 °C. The samples were kept for 16–24 h in hydrochloric acid solution (pH 1) to dissolve possible contaminants. The scale bars: (a) and (b) – 1 ?m and (c) – 10 ?m.

Now a series of interesting micrographs:



The caption:

Figure 8. SEM images of silicon structures obtained by SiO2 electro-deoxidation using a Mo wire contacting electrode at 1.10 V vs Ca2+/Ca for 1 h in molten CaCl2 at 850 °C. (a) Cross section shows the columnar array (A) with the columns formed perpendicular to the SiO2 precursor (C) and a partly deoxidized interface with the oxygen concentration 65.0 at. % (B). (b) Hexagonal Si columns obtained at the given conditions. Reprinted with permission from ref (33). Copyright 2005 The Electrochemical Society.

The caption:

Figure 9. SEM images of the surface obtained when deoxidizing the amorphous SiO2 layer on p-Si ⟨100⟩ in molten CaCl at 900 °C for 20 min. Thickness of the silica layer was 0.3 ?m (a)–(c) and 2 ?m (d). Polarization vs graphite pseudoreference electrode: ?0.3 V (a), 0.6 V (b), ?1.1 V (c), and ?0.9 V (d). The scale bars are 1 ?m for all images.

The caption:

Figure 10. SEM view of the Si fragments on AISI 304 stainless steel after 4 h exposure at 900 °C to oxygen. The Si film (10 ?m) was deposited on the substrate by radio frequency magnetron sputtering. The temperature has been increased from ambient to 900 °C at the rate 5 °C/min. The scale bar is 400 ?m.

The caption:

Figure 11. Pristine pellet prepared by the pressing of SiO2 powder and sintering at 900 °C (b) and that after electrolysis at ?1.2 V vs the Pt reference electrode in molten CaCl2 at 900 °C (a). FESEM image (c) shows the developed nanowires. The XRD pattern (d) characterizes the crystalline structure of nanowires and the amorphous structure of the precursor. The EDS spectrum (e) identifies some oxygen present on the nanowires. Reprinted with permission from ref (27). Copyright 2009 The Royal Society of Chemistry.

The paper also contains some interesting remarks on silicon carbide, which I will point out to my son.

There is also a discussion of that much loved, much hyped, but actually wasteful topic of battery electrodes for batteries.

Some concluding remarks from the authors, wherein reference to the calcium ruthenate/calcium titanate system is made:

Silicon electrochemistry in molten salts is a rapidly growing field, which offers environmentally friendly and energy efficient principles of silicon production and processing for sustainable energy generation and storage. This field is of foremost interest owing to its potential to significantly contribute to the low-carbon economy.

Considerable attention attracted electrochemical SiO2 reduction to produce SoG-Si with a negligible carbon footprint. Silicon can be electrochemically extracted from silica without dissolving it in a molten salt electrolyte. The process opens an environmentally friendly and energy efficient route for electrochemical production of pure silicon. An advantage of the electrochemical method includes the use of electrons as clean reduction agents instead of harmful chemicals, as well as simplicity owing to one-step processes.
Carbon-based anodes were predominantly used for silica decomposition. Application of such anodes, however, is detrimental in several aspects: (i) anodic emission of undesirable COx; (ii) cathodic reduction of COx with production of carbon and carbides, which contaminates the product as well as reduces output of Si extraction; and (iii) when operating on a large scale, the anode is gradually consumed, and the eroded carbon particles can shortcut the cell. A challenging goal is the creation of anode materials, which could sustain high temperatures and operate without greenhouse-gas emission. Great potential showed solid solution of CaRuO3 and CaTiO3; however, the high cost of such an electrode limits its wider application...

Derek Fray, is, in my opinion, a great man, whose ideas will surely change the world, even if it is true that electricity is an inherently thermodynamically degraded form of energy - there are some things that electricity does that nothing else can do as well, and even as cleanly, depending on how the electricity is generated.

This said, my respect for Dr. Fray being high, I am a skeptic on the widely belief that silicon, carbon's cogener, will save the world. It hasn't; after decade upon decade of wild cheering and the expenditure of trillions of dollars, it isn't; and my studied opinion it won't.

Silicon is a readily available and it is, to be sure, an extremely useful element. This computer on which I write depends on it. Refractory silicon compounds represent an important technology in materials science. Our building are structured from silicates, it lines our beaches. Still, no, so called "renewable energy" based on silicon isn't and never will be "green," if only because of its extraordinarily low energy to mass ratio.

I greatly admired the wonderful Senator Warren's interview with Rachel Maddow which reflects Senator Warren's magnificent intellect, in which she said - referring to a path to addressing the goal of restraining the nasty stuff that flies around in political campaigns - (I paraphrase) that if something doesn't work, you choose another path.

This attitude, despite my profound disagreement with her on her "plan" to address climate change, is why I placed her near or at the top of the list of the Democratic Candidates who should displace the orange criminal fool.

I will tell you that her stated plan for energy during the campaign, nearly identical with the person most closely identified with her in popular imaginations - although he, Senator Sanders, is nowhere her intellectual equal and never will be as he is a fool - will not work, no matter how fond we on the left are of this tired and discredited idea that so called "renewable energy" will save the world.

It won't. There is a reason so called "renewable energy" was abandoned in the 19th century.

Again, Dr. Fray is a great man, but in the spirit of the recently passed Freeman Dyson, we need some heresy with respect not to his skepticism about climate change itself, but with respect to addressing it.

This morning I attended a lecture by Professor Asa Rennermalm who is studying the Greenland Ice Sheet.

What we are doing is not working. I hope we can all agree to choose another path, if not now, as soon as possible because, as I often say:

"History will not forgive us; nor should it."

Enjoy the rest of the weekend.

[Warpy]

1. Well, any comparison with standard nuclear reactors

via instantaneous deaths is unfair. Radiation kills slowly. Things that go "bang" kill quickly. People need to pick their poison, slow and undetectable or loud and dramatic.

The main reason what we are doing isn't working is structural. There is so much exciting research out there that simply lacks the opportunity to scale up because of extreme risk aversion.

[NNadir]

2. The real issue is the number of deaths per unit of energy as well as the concept of DALYs

I often joke with my wife we she complains about growing old that the alternative to growing old was dying young.

Now, we will all die. Is it better for someone to die in an explosion at the age of 30 or at the age of 70 (when one might have lived to be 80) from long term exposure to radiation?

Epidemiologists attempt to define these terms in units called "DALY" or Disability Adjusted Life Years. It is not a question of whether your ultimate loss of life has radiation, or air pollution, or chemical exposure effects contributing to it; but rather a question of how long your life actually is, and whether you can be functional and productive in that period.

For example:

The tennis player Maria Sharapova's mother was pregnant with her in the Belarus region when the Chernobyl reactor blew up. I've seen this covered many times in the media. She nevertheless went on to an athletic career, an exceptional one. Now, most of the cancers associated with the release of Chernobyl will take place in people who were children at the time, specifically in the form of thyroid cancer. It is therefore true that she has a higher risk of thyroid cancer than say, you. (I have a higher risk than you do, since I used to make I-125 radiolabeled RIA kits when I was young.) Has she (and I) have less or more of a life than the guys blown up in a silane plant?

Suppose the workers in the plant averaged 40 years of age. If life expectancy in Japan is 80 years - I don't actually know the number - one can assume that the DALY is 5*40 = 200 years.

Now, it is very probable that my life may be shortened by exposure to radiation, air pollution, nasty chemicals with which I've worked - and I have worked with some very "deadly" chemicals although I'm obviously still alive - but frankly, I don't have a lot of DALYs left to contribute since I'm an old short fat bald guy. It is most likely that any DALYs I lose will be insignificant because despite all the risks I have faced and experienced, I don't have that many years to live before reaching average life expectancy. Probably what will kill me is being fat.

I used to work extensively - after working with radiation - with the war gas phosgene when I was in my early thirties. If I had dropped a flask with a few kilos of phosgene solutions in it, it is possible that I would have died at 30, in which case, if life expectancy is 76 years, the DALYs associated with my death would have been 46 years. But if on average, 5000 people worked with phosgene and the average life span of those people was 75 years, my particular case would not prove that phosgene is deadly, any more than Maria Sharapova's case proves that Chernobyl was harmless which it was not.

Nevertheless, solar energy is a trivial form of energy. For more than three decades the amount of energy produced by the solar industry despite vast choruses singing praises of it, is laughably trivial compared to the amount of energy produced by nuclear energy despite vast armies of critics carrying on about how "dangerous" it is, in my view, a tragic army of people wallowing in selective attention who are, quite clearly, functional idiots.

One wonders if the solar industry were able to produce as much energy as the nuclear energy has been routinely producing since 1990 while put on hold while air pollution killed something like 200,000,000 people in the period in which nuclear power was being derided as "too dangerous," roughly 28 exajoules per year, how many lives would be lost to the challenging chemistry associated with polysilicon preparation. I have to believe that we'd have a large number of people carrying on about dangerous solar energy.

The most dangerous form of energy on this planet in this century is also the fastest growing in units of energy, not as is often illiterately asserted in units of peak power, is dangerous fossil fuels, led by coal, even though we keep hearing that coal is dead.

According to the Global Burden of Disease report, Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015 (Lancet 2016; 388: 1659–724) in the year 2015, the most probable number of DALYs attributed to air pollution was 167,290,000. This number represents the most probable value in 2015 of the number of people who died from air pollution was 6,485,000. (The range of probable error is given in the table on page 1677.) Dividing the DALY by deaths gives an estimate of how many years of life the average person who died from air pollution related causes in 2015 was 25 years, an unsurprising number since it is recognized that children under the age of 5 are particularly susceptible to air pollution related deaths.

And this figure gets at the real issue behind why the solar industry is dangerous. It is not dangerous because it kills people often like those poor guys in the Mitsubishi trichlorosilane plant. The reason it is dangerous is because it is totally ineffective at preventing air pollution deaths. Despite trillion dollar sums being thrown at it; despite vast armies of scientists, engineers and funding agencies trying to make it work to appeal to popular mythology, it hasn't worked; it isn't working and there is no evidence that it will work. And yet we carry on, repeating the same experiment over and over and over and over and over expecting a different result. Meanwhile six to seven million people die each year, every year, from air pollution.

The Lancet article is open sourced. You should feel free to look into it to find about all those radiation deaths upon which everyone is always focusing, a focus that actually kills people because nuclear energy, despite its risks, saves lives.

Nuclear energy need not be risk free; it need not be harmless to be vastly superior to everything else. It only needs to be vastly superior to everything else, which it is.

Thanks for giving me the opportunity to explore this issue.

Have a nice Sunday.